Request a
30 day Trial or
Demonstration
Impala Multiple-point Statistics Facies Modeling
MPS is rapidly becoming the tool of choice for facies modeling in the E&P industry, and has significant advantages over object-based and 2-point voxel-based modeling techniques. MPS simulations can be effectively conditioned to external data in the form of wells, 3D facies proportion and depositional direction data, and at the same time produce complex and laterally persistent patterns as observed in outcrops and present day analogues.
MPS facies modeling involves two steps: first a 2D or 3D ‘training image’ is analyzed for facies patterns occurring, secondly the registered patterns are used to compose a simulation model, honoring external data. A clear benefit to geological users is that valid MPS models can be generated without detailed knowledge of the underlying math and statistical assumptions, concentrating more on geological concepts.
Conventional MPS simulations are computation and memory intensive. The performance of Impala, however, is greatly optimized by the implementation of two important features:
- Memory usage and search efficiency is optimized using a list approach rather than search trees.
- Impala is fully parallelized, making efficient use of multiple processors and clusters.
In the Impala module of JewelSuite, training images are separate 3D grids that are detached from the actual simulation model. This set-up facilitates easy exchange of training images between projects, e.g. import from a library of training images for various depositional geometries.
The implementation of Impala in JewelSuite gives significant additional benefits:
- Workflow guidance: The Impala plug-in interface provides detailed access for the expert and tested defaults for newer users.
- Efficiency: common controls like that for depositional gradient are wrapped into intuitive GUI’s and powerful automated methods.
- Ease of Use: extensive Graphical Editor options make it easy and even fun to build 2D/3D training models.